Molecular transition metal catalysts offer unique potential for the production of fine chemicals. Chemical processes carried out in the presence of well defined molecular catalysts often only require mild, easily accessible conditions, fewer sacrificial reagents, and can selectively produce a desired product with minimal waste. The active site of a transition metal catalyst can be varied by the use of a hybrid ligand, which employs a combination of groups with different binding affinities for the metal center. Hybrid ligands possessing both substitutionally inert and labile donors, called “hemilabile” ligands, offer an added dimension to catalysis since the weakly binding donor can be displaced from the metal center by a substrate to facilitate the chemical transformation. However this labile donor, in conjunction with an inert donor, can also offer chelate stabilization of the catalyst in the event of coordinative unsaturation at the metal center, a feature which can serve to enhance catalyst longevity. A major goal of the research reported herein is to understand the mechanisms by which hemilabile processes occur within ortho-phosphinoaniline complexes of rhodium and ruthenium and, in turn, how such features might affect catalytic characteristics. To this end, a comparison of catalytic activities of related hemi- and non-labile complexes has been carried out. The ability for two metal atoms held in close proximity to have a cooperative effect on substrate activation or catalysis has also inspired the generation of a series of binuclear compounds bridged by bis(ortho-phosphinoaniline) ligands. In addition to hemilabile and catalytic features, many unique ligand geometries and coordination modes are also observed, particularly by altering the substituents on labile amine donors. Non-labile complexes can also be prepared by deprotonation of labile amine donors to produce ortho-phosphinoanilido species, which display reactivity patterns and structural features distinct from the those of their hemilabile congeners. The amido complexes, which are effective toward ketone transfer hydrogenation and olefin silylation reactions, display interesting features, and in the first case, the possibility of a reaction mechanism unprecedented for transition metal catalysts is discussed. Evidence supporting the operation of such an unexpected mechanism could have important implications for the design and operation of new and more effective transition metal catalysts. / Chemistry

Depuis quelques années, la recherche et le développement dans le domaine des matériaux électroluminescents suscite de plus en plus d’intérêt. La crise énergétique que nous connaissons actuellement motive un changement de mentalité de la société, de plus en plus soucieuse d’économiser l’énergie. Peu à peu, la commercialisation de diodes électroluminescentes telles que les LEDs (Light-Emitting Diodes) ou les OLEDs (Organic Light-Emitting Diodes) s’est implantée et elle connait aujourd’hui un essor considérable. Néanmoins, de nombreuses améliorations doivent encore être apportées afin d’en optimiser les propriétés. C’est dans cette thématique des matériaux électroluminescents que s’est focalisé notre travail. La préparation de nouveaux composés et l’étude de leurs propriétés photophysiques représentent en effet des étapes déterminantes et indispensables à la fabrication de ces dispositifs. Notre travail s’est concentré plus particulièrement sur des complexes de Rh(III), Ir(III) et Ru(II) incorporant un ou plusieurs dérivés du dipyrrométhène. Le premier volet de cette thèse est consacré à la synthèse de quatre complexes mononucléaires [Rh(dipy)3], [Rh(ppy)2dipy], [Ir(ppy)2dipy], [Ru(phen)2dipy]+ dont nous avons ensuite étudié les propriétés photophysiques et électrochimiques. Des calculs théoriques TD-DFT ont également permis de conforter l’attribution de la nature des différents états excités responsables de la luminescence de ces composés. Le second volet de cette thèse est consacré à la mise au point de complexes dinucléaires à l’aide d’un dérivé pontant du dipyrrométhène, avec lequel deux composés ont été obtenus : [Ir2(ppy)4µdipy] et [Rh2(ppy)4µdipy]. Une étude complète du composé d’iridium nous a permis de déterminer ses propriétés. Les nombreux problèmes de solubilité que nous avons rencontrés avec le composé de rhodium ne nous ont pas permis de mener des études approfondies de ce composé. Lors d’un essai de synthèse du composé dinucléaire d’iridium, nous avons isolé le complexe mononucléaire pontant [Ir(ppy)2µdipyH]. Afin de l’engager dans la préparation d’un complexe hétérodinucléaire de type Bodipy, nous avons préalablement mis au point la synthèse d’un bodipy mononucléaire et nous avons examiné ses propriétés optiques. Nous nous sommes dès lors basé sur les conditions de synthèse de ce bodipy mononucléaire pour tenter de réaliser, sans succès, la synthèse du complexe hétéronucléaire [Ir(ppy)2µdipyBF2]. L’ensemble des résultats obtenus au cours de ce travail, nous ont permis de comprendre l’effet du métal de transition et celui de la dérivatisation du dipyrrométhène sur les propriétés photophysiques des complexes formés. Il serait dès lors intéressant de tirer parti de l’ensemble de ces résultats pour mettre au point un nouveau dérivé pontant du dipyrrométhène afin d’optimiser les propriétés des complexes formés, et à terme permettre la préparation de réseaux organométalliques performants et exploitables dans des dispositifs électroluminescents.

This thesis describes the discovery of catalytic reactions that create carbon-carbon bonds stereoselectively between substrates bearing an alkene and organoboronic acids reagents. Chiral rhodium(I) catalysts were found to react with various meso-symmetrical substrates, thereby resulting in enantioselective desymmetrization reactions. The methodologies presented herein allow the rapid synthesis of several chiral functionalized molecules; including branched homoallylic alcohols, cyclopentenyl hydrazines, and ketohydrazines. The thesis is divided according to three main transformations: asymmetric allylic substitution of allylic carbonates, asymmetric ring-opening of [2.2.1]-diazabicyles, and carbonylation of alkenes or alkynes. Chapter 2 details the investigations of a ligand-controlled catalytic process to prepare either trans-2-arylcyclopent-3-enols (up to 94% ee), or trans-4-arylcyclopent-2-enols (up to 99% ee) as the major products starting from cyclic meso allylic dicarbonates. This rhodium-catalyzed methodology was extended to include linear allylic dicarbonates, thereby yielding chiral 2-arylbut-3-enols with up to 95% ee. An enantioselective desymmetrization of strained alkenes by ring-opening of meso bicyclic hydrazines is described in Chapter 3. The reaction allows one to prepare trans-2-arylcyclopent-3-enyl hydrazides with up to 99% ee. In addition, an enantioselective hydroarylation process was identified to yield 5-aryl-2,3-diazabicyclo[2.2.1]heptanes. Mechanistic investigations showed that the reaction proceeds via an unusual C-H activation/1,4-migration of the rhodium catalyst. Finally, Chapter 4 outlines the development of a mild catalytic acylation of pi systems. This mode of reactivity was optimized to promote the desymmetrization of [2.2.1]-diazabicycles via a formal allylic substitution with acyl anions as nucleophiles. The method yields densely functionalized trans-2-ketocyclopent-3-enyl hydrazides. In addition, preliminary studies demonstrate that the rhodium(I)-catalyzed acyl anion addition is also possible with other pi electrophiles. For example, with alkyne, it provided a synthesis of cyclopentenones that complements the Pauson-Khand reaction. Overall, the catalytic transformations reported herein give access to seven classes of products stereoselectively; starting from simple reagents.

We report the application of our rhodium-catalyzed nucleophilic ring-opening methodology to the enantioselective synthesis of nitrogen-substituted polycyclic heterocycles. By using a cationic Rh(I) triflate catalyst in the presence of the chiral Josiphos ligand PPF-PtBu2, the ring opening reactions on dihydrooxaquinoline and dihydrooxaisoquinoline using different nucleophiles afford access to multiple dihydroquinolines and dihydroisoquinolones in high yield and high enantioselectivity (up to 99% total yield and >99%ee). A variety of nucleophiles were shown to be compatible with the catalytic system. The electronic effects in the new ring opening reactions were investigated using a variety of nucleophiles. It was found that reactivity and enantioselectivity of the ring opening products depends on the electronic effects as well as the position of the substituents on the substrates. Good yields and high ee of regioisomeric products are obtained using electron donating substituents, whereas electron withdrawing substituents decelerate the reactions.

We report the application of our rhodium-catalyzed nucleophilic ring-opening methodology to the enantioselective synthesis of nitrogen-substituted polycyclic heterocycles. By using a cationic Rh(I) triflate catalyst in the presence of the chiral Josiphos ligand PPF-PtBu2, the ring opening reactions on dihydrooxaquinoline and dihydrooxaisoquinoline using different nucleophiles afford access to multiple dihydroquinolines and dihydroisoquinolones in high yield and high enantioselectivity (up to 99% total yield and >99%ee). A variety of nucleophiles were shown to be compatible with the catalytic system. The electronic effects in the new ring opening reactions were investigated using a variety of nucleophiles. It was found that reactivity and enantioselectivity of the ring opening products depends on the electronic effects as well as the position of the substituents on the substrates. Good yields and high ee of regioisomeric products are obtained using electron donating substituents, whereas electron withdrawing substituents decelerate the reactions.

This thesis describes the discovery of catalytic reactions that create carbon-carbon bonds stereoselectively between substrates bearing an alkene and organoboronic acids reagents. Chiral rhodium(I) catalysts were found to react with various meso-symmetrical substrates, thereby resulting in enantioselective desymmetrization reactions. The methodologies presented herein allow the rapid synthesis of several chiral functionalized molecules; including branched homoallylic alcohols, cyclopentenyl hydrazines, and ketohydrazines. The thesis is divided according to three main transformations: asymmetric allylic substitution of allylic carbonates, asymmetric ring-opening of [2.2.1]-diazabicyles, and carbonylation of alkenes or alkynes. Chapter 2 details the investigations of a ligand-controlled catalytic process to prepare either trans-2-arylcyclopent-3-enols (up to 94% ee), or trans-4-arylcyclopent-2-enols (up to 99% ee) as the major products starting from cyclic meso allylic dicarbonates. This rhodium-catalyzed methodology was extended to include linear allylic dicarbonates, thereby yielding chiral 2-arylbut-3-enols with up to 95% ee. An enantioselective desymmetrization of strained alkenes by ring-opening of meso bicyclic hydrazines is described in Chapter 3. The reaction allows one to prepare trans-2-arylcyclopent-3-enyl hydrazides with up to 99% ee. In addition, an enantioselective hydroarylation process was identified to yield 5-aryl-2,3-diazabicyclo[2.2.1]heptanes. Mechanistic investigations showed that the reaction proceeds via an unusual C-H activation/1,4-migration of the rhodium catalyst. Finally, Chapter 4 outlines the development of a mild catalytic acylation of pi systems. This mode of reactivity was optimized to promote the desymmetrization of [2.2.1]-diazabicycles via a formal allylic substitution with acyl anions as nucleophiles. The method yields densely functionalized trans-2-ketocyclopent-3-enyl hydrazides. In addition, preliminary studies demonstrate that the rhodium(I)-catalyzed acyl anion addition is also possible with other pi electrophiles. For example, with alkyne, it provided a synthesis of cyclopentenones that complements the Pauson-Khand reaction. Overall, the catalytic transformations reported herein give access to seven classes of products stereoselectively; starting from simple reagents.

The [2+2+2] cycloaddition reaction involves the formation of three carbon-carbon bonds in one single step using alkynes, alkenes, nitriles, carbonyls and other unsaturated reagents as reactants. This is one of the most elegant methods for the construction of polycyclic aromatic compounds and heteroaromatic, which have important academic and industrial uses. The thesis is divided into ten chapters including six related publications. The first study based on the Wilkinson’s catalyst, RhCl(PPh3)3, compares the reaction mechanism of the [2+2+2] cycloaddition process of acetylene with the cycloaddition obtained for the model of the complex, RhCl(PH3)3. In an attempt to reduce computational costs in DFT studies, this research project aimed to substitute PPh3 ligands for PH3, despite the electronic and steric effects produced by PPh3 ligands being significantly different to those created by PH3 ones. In this first study, detailed theoretical calculations were performed to determine the reaction mechanism of the two complexes. Despite some differences being detected, it was found that modelling PPh3 by PH3 in the catalyst helps to reduce the computational cost significantly while at the same time providing qualitatively acceptable results. Taking into account the results obtained in this earlier study, the model of the Wilkinson’s catalyst, RhCl(PH3)3, was applied to study different [2+2+2] cycloaddition reactions with unsaturated systems conducted in the laboratory. Our research group found that in the case of totally closed systems, specifically 15- and 25-membered azamacrocycles can afford benzenic compounds, except in the case of 20-membered azamacrocycle (20-MAA) which was inactive with the Wilkinson’s catalyst. In this study, theoretical calculations allowed to determine the origin of the different reactivity of the 20-MAA, where it was found that the activation barrier of the oxidative addition of two alkynes is higher than those obtained for the 15- and 25-membered macrocycles. This barrier was attributed primarily to the interaction energy, which corresponds to the energy that is released when the two deformed reagents interact in the transition state. The main factor that helped to provide an explanation to the different reactivity observed was that the 20-MAA had a more stable and delocalized HOMO orbital in the oxidative addition step. Moreover, we observed that the formation of a strained ten-membered ring during the cycloaddition of 20-MAA presents significant steric hindrance. Furthermore, in Chapter 5, an electrochemical study is presented in collaboration with Prof. Anny Jutand from Paris. This work allowed studying the main steps of the catalytic cycle of the [2+2+2] cycloaddition reaction between diynes with a monoalkyne. First kinetic data were obtained of the [2+2+2] cycloaddition process catalyzed by the Wilkinson’s catalyst, where it was observed that the rate-determining step of the reaction can change depending on the structure of the starting reagents. In the case of the [2+2+2] cycloaddition reaction involving two alkynes and one alkene in the same molecule (enediynes), it is well known that the oxidative coupling may occur between two alkynes giving the corresponding metallacyclopentadiene, or between one alkyne and the alkene affording the metallacyclopentene complex. Wilkinson’s model was used in DFT calculations to analyze the different factors that may influence in the reaction mechanism. Here it was observed that the cyclic enediynes always prefer the oxidative coupling between two alkynes moieties, while the acyclic cases have different preferences depending on the linker and the substituents used in the alkynes. Moreover, the Wilkinson’s model was used to explain the experimental results achieved in Chapter 7 where the [2+2+2] cycloaddition reaction of enediynes is studied varying the position of the double bond in the starting reagent. It was observed that enediynes type yne-ene-yne preferred the standard [2+2+2] cycloaddition reaction, while enediynes type yne-yne-ene suffered β-hydride elimination followed a reductive elimination of Wilkinson’s catalyst giving cyclohexadiene compounds, which are isomers from those that would be obtained through standard [2+2+2] cycloaddition reactions. Finally, the last chapter of this thesis is based on the use of DFT calculations to determine the reaction mechanism when the macrocycles are treated with transition metals that are inactive to the [2+2+2] cycloaddition reaction, but which are thermally active leading to new polycyclic compounds. Thus, a domino process was described combining an ene reaction and a Diels-Alder cycloaddition. / La reacció de cicloaddició consisteix en la formació de tres enllaços carboni-carboni en un únic pas de reacció on poden estar involucrats alquins, alquens, nitrils, carbonils i altres compostos insaturats. És un dels mètodes més elegants per a la construcció de compostos aromàtics i heteroaromatics policíclics amb importants usos acadèmics i industrials. La tesi es divideix en deu capítols que contenen sis publicacions relacionades. El primer estudi es basa en el catalitzador de Wilkinson, RhCl(PPh3)3, on es compara el mecanisme de reacció del procés de cicloaddició d’acetilè pel que s’obté amb el model del complex, RhCl(PH3)3. Aquest projecte de recerca va ser iniciat per estudiar la substitució de lligands PPh3 per PH3 en els estudis de DFT, que s’aplica habitualment per reduir el cost computacional, tot i que els efectes electrònics i estèrics produïts pels lligands PPh3 són molt diferents dels creats per PH3. Malgrat algunes diferències observades, es va constatar que la substitució de PPh3 per PH3 en el catalitzador pot ser utilitzada per reduir el cost computacional de manera significativa i a l’hora obtenir resultats qualitativament acceptables. Un cop obtinguts els resultats anteriors, es va utilitzar el model del catalitzador de Wilkinson, RhCl(PH3)3, per l’estudi teòric de diferents reaccions de cicloaddició amb sistemes insaturats duts a terme al laboratori. En el grup de recerca es va trobar que en el cas de sistemes totalment tancats, concretament els macrocicles de 15 i 25 baules, poden donar sistemes benzènics policíclics excepte en el cas del macrocicle de 20 baules, que va resultar inactiu vers el catalitzador de Wilkinson. En aquest estudi, la realització de càlculs teòrics va permetre determinar l’origen de la diferent reactivitat del macrocicle de 20 baules, on es va trobar que la barrera d’activació de l’addició oxidativa entre dos alquins és molt més alta que les que es van obtenir pel macrocicle de 15 i 25 baules. Aquesta barrera es va atribuir bàsicament a l’energia d’interacció, la qual correspon a l’energia que s’allibera quan els dos reactius deformats interaccionen en l’estat de transició. Concretament el principal factor que hi contribueix és que el macrocicle de 20 baules presenta més estabilitat i més deslocalització de l’orbital HOMO en el pas d’addició oxidativa. A més a més, es va observar que la formació d’anells de 10 baules durant la cicloaddició del macrocicle de 20 baules presenta impediments estèrics importants. Per altra banda, en el Capítol 5 es presenten estudis electroquímics realitzats en col•laboració amb la Prof. Anny Jutand de París, que van permetre estudiar el cicle catalític de la reacció de cicloadició entre un dií i un monoalquí. Es van obtenir així les primeres dades cinètiques dels dos principals passos del cicle catalític amb el complex de Wilkinson, on es va observar que el pas determinant de la reacció pot variar en funció de l’estructura dels reactius de partida. En el cas en què en la reacció de cicloaddició participin dos triples i un doble enllaç en la mateixa molècula (endiins), és conegut que l’addició oxidativa pot donar-se entre dos triples enllaços o un triple i un doble enllaç. El model del catalitzador de Wilkinson va ser utilitzat mitjançant càlculs DFT per analitzar els diferents factors que poden influir en el mecanisme de reacció. Aquí es va observar que els endiins cíclics sempre prefereixen l’addició oxidativa entre els dos triples enllaços, mentre que els acíclics tenen diferent preferència en funció del linker i dels substituents presents en els triples enllaços. A més a més, el mateix model de Wilkinson es va utilitzar per explicar els resultats experimentals realitzats en el Capítol 7 on s’estudia la reacció de cicloaddició d’endiins variant la posició del doble enllaç en el reactiu de partida. Es va observar que els sistemes in-en-in preferien la cicloaddició convencional donant el producte ciclohexadienic esperat, mentre que els sistemes in-in-en patien una β-eliminació seguida d’una eliminació reductiva del catalitzador de Wilkinson i donant, finalment, productes ciclohexadiènics els quals són isòmers dels que s’obtindrien mitjançant una cicloaddició convencional. Finalment, l’últim capítol d’aquesta tesi es basa en l’ús de càlculs DFT per determinar el mecanisme de reacció quan els macrocicles són tractats amb metalls de transició inactius per donar la reacció de cicloaddició, però reaccionen tèrmicament obtenint nous compostos policíclics. Així es va descriure un procés dòmino on es combina una reacció ene seguida d’una cicloaddició de Diels-Alder.

Pt-group metals were some of the first metals to be studied as catalysts for industrial use. The goal of these studies was to ascertain a fundamental understanding of CO oxidation and acetylene cyclotrimerization reactions on Ptgroup metals. A further goal was to determine the optimal conditions for each reaction. CO oxidation on Rh(111),Pt(100), and Pd(100) was scrutinized on various oxide surfaces from chemisorbed to bulk metal oxides. Low pressure reactions on Rh(111) reveal the highest activity was a CO uninhibited surface with <1ML of chemisorbed oxygen. Pt(100) high pressure oxidation revealed that only <1ML oxygen is formed during high pressures reactions. High pressure CO oxidation reactions on Pd(100) show oxygen penetration after CO has been consumed; however, during the highest activity XPS found only chemisorbed species. The cyclotrimerization of acetylene to benzene is another reaction found in industry typically carried out on Pd. The active site is considered to be a 7 atom configuration with 6 atoms surrounding a central atom. By adding relatively catalytically inert Au atoms to the active Pd(111) surface the acetylene coupling activity is enhanced. Cyclization activity is a function of the surface composition and the surface structure. A single Pd atom surrounded by six Au atoms is found to have the highest activity at 300K for acetylene cyclotrimerization.

Thesis (Ph. D.)--University of Washington, 2000. / Vita. Includes bibliographical references (leaves 136-144).

Halogenated hydrophobic organic compounds (HHOCs) such as 1,2,4,5-tetrachlorobenzene (TeCB) present a threat to both human health and the environment. The common occurrence and recalcitrant nature of HHOCs as soil contaminants necessitate an effective soil remediation method. Wee and Cunningham (2008, 2011, 2013) proposed a clean-up technology called Remedial Extraction and Catalytic Hydrodehalogenation (REACH), which pairs solvent extraction of HHOC contaminants from soil with catalytic hydrodehalogenation to destroy contaminants. Wee and Cunningham (2008, 2011, 2013) utilized a palladium (Pd) catalyst to hydrodehalogenate TeCB to benzene. However, benzene is still a toxic contaminant. Prior research has demonstrated that Pd-catalyzed hydrodehalogenation (HDH) can be paired with Rh-catalyzed hydrogenation to transform TeCB to cyclohexane, which is a less toxic end product (Osborn 2011; Ticknor 2012). However, there remains a need to quantify the effects of different operating conditions on the catalytic reaction rates upon which the technology relies. It was hypothesized that (1) an increased ratio of water to ethanol in water/ethanol solvents would increase the reaction rates of both Pd-catalyzed HDH and Rh-catalyzed hydrogenation, and (2) catalytic reaction rates would be constant above a hydrogen pressure threshold, but would decrease with decreasing hydrogen pressure beneath the threshold. Thus, the objective of this thesis was to contribute to the development of optimal operating parameters for the REACH technology by quantifying the effects of solvent composition and hydrogen pressure on the catalytic conversion of TeCB to cyclohexane in water/ethanol solvents in a batch reactor. Complete conversion of TeCB to cyclohexane was achieved at all experimental conditions tested. The data were consistent with an apparent first-order kinetics model where Pd-catalyzed HDH and Rh-catalyzed hydrogenation occur in series. The effects of three water/ethanol solvent compositions (33:67, 50:50, 67:33) were investigated at 50 psi hydrogen pressure. HDH rate coefficients increased monotonically with an increasing fraction of water in the solvent. When the water fraction in the solvent was increased from 50% to 67%, a larger HDH rate coefficient increase was observed than when the water fraction was increased from 33% to 50%. In both cases, the observed increases were statistically significant at a 95% confidence level. For hydrogenation, rate coefficients at 33% and 50% water were approximately equal. The hydrogenation rate coefficient at 67% water was much greater than the rate coefficients at 50% and 33% water, but the increase was not statistically significant at a 95% confidence level. The observed time for complete conversion of TeCB to cyclohexane decreased with an increasing fraction of water in the solvent, from 12-18 hours with a 33% water solvent to 8-12 hours with a 50% water solvent, and to 1-1.5 hours with a 67% water solvent. The effects of three hydrogen pressures (50 psi, 30 psi, 10 psi) were investigated with a 50:50 water/ethanol solvent. HDH rate coefficients increased monotonically with decreasing hydrogen pressure, though the trend was not statistically significant at a 95% confidence level until the pressure was decreased from 30 psi to 10 psi. This trend can be attributed to the displacement of TeCB by hydrogen on the catalyst surface at higher hydrogen pressures. For hydrogenation, the data suggest that rate coefficients are independent of hydrogen pressure in the pressure range of 10-50 psi, since no statistically significant hydrogen pressure effect was observed. Complete conversion of TeCB to cyclohexane was achieved at hydrogen pressures as low as 5 psi. These findings suggest that a greater fraction of water in the solvent should be utilized in the REACH system when feasible to maximize catalytic reaction rates. These findings also suggest that the REACH system could be operated at hydrogen pressures as low as 5 psi, which would further improve the safety of the technology.